47. Diversity and the Latin language

This post is also available in: Catalan

James Adams
All Souls College Oxford


Those learning Latin grammar for the first time may be tempted to look upon the language as a fossilised thing, rigidly standardised. And yet it was to evolve in the different parts of the Roman Empire into a variety of different (Romance) languages, closely related to one another but different enough that a native speaker of one will usually have to learn how to speak another. Latin was once a living language, evolving gradually. In the last forty years or so much has been learnt about the diversity of Latin during the Roman period, thanks particularly to the ongoing discovery of writing tablets from different parts of the former Empire. These reflect not the usage of high literature but that of ordinary people, who in some cases were dictating to scribes, so that we may sometimes be observing specimens of mundane speech. Tablets have been found in various parts of the Empire, with Britain a particular source of new discoveries. Many tablets come from the Roman military base, Vindolanda, on Hadrian’s Wall, dating roughly from the early second century AD. These are often private letters. A different category consists of curses directed against someone who has wronged the writer. These, sometimes a substitute for a police service that did not exist, have been turning up over a long period from all over the ancient world in Greek and Latin, but substantial Latin discoveries have been made in Britain, at for example Bath and Uley in Gloucestershire.

One of the Vindolanda letters (Tab. Vindol. 291) has possibly the earliest piece of handwriting by a woman extant. Claudia Severa invites her friend Lepidina to her birthday party. The letter is written in two hands, one for the formal invitation, and the other for the endearment addressed to Lepidina at the end. The invitation was written by a scribe, and then Claudia took over and closed with her affectionate greeting, calling Lepidina anima mea. . . . karissima, ‘my dearest soul’. Lepidina uses the spelling with k for carissima, which was recommended by grammarians before the letter a. The change of hands is not uncommon in such letters.

The regional and social diversity of the Latin language receives comment from as early as about 200 BC. For instance, the rustic character Truculentus in Plautus’ play of that name uses the word rabonem for arrabonem, which is picked up at once by another speaker and described as a ‘monster’ (beluam). Truculentus defends himself by citing as a (supposed) parallel a regional term ‘as used by the Praenestines’. It was no doubt thought to be funny that before a Roman audience he justified a linguistic abnormality by citing a usage from Praeneste. In the late Republic the recognition that there were regional varieties of the language outside the city began to generate an ideological debate, with some city ‘purists’ damning the ‘harshness’ of rural varieties (see e.g. Cicero, De oratore 3.42) and attempting to stamp out their influence in the city. Cicero in particular pronounces on the merits of Roman Latin, which he thought to be under threat because of the influx of outsiders (Brutus 258). Some comment on ‘rustic’ Latin was however more neutral, consisting of phonetic observations. Cicero’s learned contemporary Varro, author of a work on the Latin language, mentions a rustic pronunciation of via, as veha. Attitudes to the variation perceived between dialects of the city and those of the country were not uniform. There were some who found rural varieties old-fashioned, and cultivated them. Cicero (De oratore 3.42) refers to L. Cotta, who took delight in the ‘rustic sound of his voice’ and thought that it reflected the speech of an earlier time.

I turn again to new writing tablets and other discoveries and some evidence they provide for aspects of the diversity of Latin.

The literary word for ‘horse’ was equus, which occurs hundreds of times in classical texts. This is a word which, despite its frequency, does not survive (except in the feminine: equa ‘mare’) in any of the Romance languages, where it is caballus that provides the term for ‘horse’, a loanword into Latin of unknown origin. Caballus is rare in Latin literature, and it tends to be in low genres or to be pejorative in tone, denoting a horse of low quality. In the Vindolanda tablets equus has not yet turned up. Remarkably, caballus is the term used instead by the military personnel stationed there (four examples so far, one in a tablet just published, in 2019). As these are army animals they are unlikely to have been of low quality. The Vindolanda tablets are perhaps the only corpus extant from the Roman world in which caballus is preferred to equus. Here is evidence for the social diversity of the language in the early second century. The man in the street used caballus, whereas high literature used equus. The everyday term remained largely submerged, but writing tablets have brought it to the surface and shown that it was not merely derogatory.

Another such case is provided by the form of the word for ‘blood’, classical Lat. sanguis, accusative case sanguinem. In writing tablets from Uley, Bath and the Hamble Estuary in Britain a modified accusative form sanguem has appeared recently five times, reflecting a standardisation of the number of syllables of the different case forms. It now becomes clear that it was this submerged form (and not the literary, neuter, word sanguen, as was previously thought) that generated Romance terms such as Italian sangue, French sang, Catalan sang and Portuguese sangue.

Or again, the verbal abstract noun vectura of classical Latin (of the action of transporting someone or something) is now attested in a Vindolanda tablet (600), not only in an assimilated form (vetura with ct > t), but also with a concrete meaning (= ‘wagon’). Here at an early date we have an anticipation of Fr. voiture and It. vettura.

New discoveries also throw light on contacts across the Roman Empire that contributed to the diversity of the language. For example, various Greek loanwords were introduced into Latin in Egypt, probably in military circles, during the Empire, and had no currency in Latin beyond that region. An example is amaxa ‘wagon’, < ἅμαχα, which is found in the ostraca from Wâdi Fawâkhir and also in a letter from the Myos Hormos Road. The word is also in the Greek ostraca from Wâdi Fawâkhir, and it had obviously found its way into Latin locally without spreading.

Moritix, a Celtic word meaning ‘sailor, seafarer’ (lit. ‘one going by sea’), in 2002 turned up in a Latin inscription from a site in Southwark, London (British Epigraphy Society Newsletter 8, 2002). Here is a word that had entered Latin in the Celtic provinces, denoting a type of trader. The latest attestation is suggestive of trading links between London and Celtic regions across the Channel.

Another striking item came to light in 1994 in a curse tablet from Brandon, Suffolk. The object stolen is referred to as popia. Popia is recognisable as a word without etymology meaning ‘ladle’. The word survives in Gallo-Romance, mainly with the meaning ‘ladle’. The attestation from Brandon again suggests a connection between Gaul and Britain. Popia must be a dialect word for ‘ladle’, as there were other terms with this meaning, such as trulla.

We do not of course depend only on writing tablets and the like for information about the linguistic diversity of Latin and its causes. Some literary evidence from the Republic was cited above for dialect variation between Rome and rural areas of Italy. I mention here just one other body of literary material, of imperial date (Christian texts), that gave an impulse to language variety. A new influence on Latin during the Empire were Bible translations. These were texts translated from Hebrew (Old Testament) and Greek (New Testament), and syntactic features were sometimes taken over into the Latin versions from the source language. With verbs of saying, for example, in classical Latin the dative case was normally used to express the addressee, whereas in the Romance languages reflexes of Lat. ad have replaced the dative, except with pronouns. Latin Bible translations seem to have been one influence giving impetus to the replacement of the dative by ad. In the Latin version of the OT ad is common with verbs of saying, under the influence of the Hebrew, and in the Gospel of John ad is also so used, under the influence of the Greek. Some Christian writers seem to have picked up this use of ad from the Vulgate, and they admitted it in their own works. Jerome in his letters was one such. It would be wrong to imply that Biblical influence was the main determinant of the switch from the dative to ad (with nouns and names), a change with a drawn-out history and complex determinants, but it played a part.

The Church fathers also made attempts to influence the language. A notable case is that of names for the days of the week. The pagan names, which alluded to pagan gods (e.g. dimarts < dies Martis in Catalan), were stigmatised, and an effort was made to introduce, after the ‘Lord’s day’ ((dies) dominica/dominicus), circumlocutions such as secunda feria ‘Monday’, tertia feria, etc. Feriae, a plural in Classical Latin, originally meant ‘festival, holy day’. The reform succeeded in Portugal (e.g. Pg. segunda-feira ‘Monday’). The circumlocutions were used in late Latin by Christian writers from other parts of the Empire too. The pilgrimage text the Peregrinatio Aetheriae, written by a woman from Gaul, makes extensive use of the circumlocutions, with ordinals from secunda to sexta. Another Gallic writer, Caesarius of Arles, in one of his published sermons urged the new names on his addressees. Despite this, the usage did not survive in Gallo-Romance.

A lexical success of Christian origin was the Greek word parabola (παραβολή), which was used in the Greek New Testament and from there as a borrowing in the Latin translations, with the meaning ‘parable, illustration’. It was to survive throughout the Romance languages with the meaning ‘word’, probably via an intermediate meaning ‘Word (of God, Christ’).

Latin thus had a diversity determined e.g. by trading contacts, army movements, the efforts of reformers, morphological simplifications, and distinctions of attitude to lexemes across different social classes, about which we are learning more form new discoveries. I have merely touched the surface above.

There is however more to diversity than regional and social variations of a single language. In a wider sense linguistic diversity is significantly diminished by imperialism and modern communications leading to language death. It has been estimated that in about 100 BC 60 different languages were spoken around the Mediterranean, whereas by AD 400 only about half a dozen of these (apart from Latin and Greek) survived. Latin had begun by eliminating the languages of Italy, and then spread further. Greek retained its high prestige and coexisted with Latin in eastern parts of the Empire. Language death is a phenomenon of widespread concern in the modern world.

We do not however hear of an aggressive Roman policy of eliminating local languages. Punic for example continued to be spoken in Africa well into the Empire. New discoveries, once again, have thrown light on local bilingualism, revealing some local languages coexisting at least for a time with Latin, and interacting with it. By far the most important evidence of this kind is provided by the records of a pottery at La Graufesenque, near Millau, France, on the left bank of the river Dourbie (published in 1988). The pottery produced Samian wares of Italian type in the style of Arretium in N. Italy. Some of the potters working there have names of Latin origin, and others have names of Gaulish origin. There had probably been immigration of potters from Arretium to South Gaul. Some of the records are in Gaulish, and some in Latin, but in others there is language switching. Latin inflections are applied to Gaulish words, and Gaulish inflections to Latin. On the whole the two languages are differentiated, but changes of language (code-switches) occur in single texts. This code-switching is consistent with a partially bilingual community in which the potters were communicating in both languages.

The diversity of Latin is revealed by various sources, but it is important to be aware of the abundant and increasing non-literary documents, which, if they come into the hands of public collections rather than private collectors, may gradually contribute to a revision of the history of the language.